In the currently discussed mechanism for many ion pumps, the access of a single binding site alternates between the two membrane surfaces, dependent on an energy-yielding reaction, while the transported ion is conducted to and from this site by specific pathways. The light-driven proton and chloride ion pumps, bacteriorhodopsin and halorhodopsin, are ideal systems to study the mechanism in which such global and local processes interact to ensure the unidirectional movement of the ion across the membrane. Although its details are specific to the two bacterial rhodopsins, the alternating access mechanism identifies principles that may apply to other ion pumps: a) control of local access through small structural changes at the active site, b) control of two alternative global protein conformations through local electrostatic effects, c) control of binding affinity and internal ion transfer through changing dielectric environments in proteins domains, d) control of specificity through subtle. differences in hydrogen bonding of the transported ion at its binding site, and e) essential involvement of the dipole properties of bound water in all processes of the ion translocation. The research strategy in this proposal is based on the similarity of the proton and chloride pumps to one another, and the recent finding that a single amino acid replacement changed the specificity of one into the other. Recombinant retinal proteins of this kind will be examined with time- resolved spectroscopy in the visible and the infrared, x-ray crystallography, and spin-label probes for properties a) through e) in order to distinguish structures and processes relevant to the binding and conduction of the transported ions from those involved in the access change of the binding sites.